CN114066786A - Infrared and visible light image fusion method based on sparsity and filter - Google Patents
Infrared and visible light image fusion method based on sparsity and filter Download PDFInfo
- Publication number
- CN114066786A CN114066786A CN202010768408.9A CN202010768408A CN114066786A CN 114066786 A CN114066786 A CN 114066786A CN 202010768408 A CN202010768408 A CN 202010768408A CN 114066786 A CN114066786 A CN 114066786A
- Authority
- CN
- China
- Prior art keywords
- image
- frequency layer
- infrared
- visible light
- images
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000007500 overflow downdraw method Methods 0.000 title claims description 13
- 238000012360 testing method Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 25
- 230000000694 effects Effects 0.000 claims abstract description 21
- 238000001914 filtration Methods 0.000 claims abstract description 12
- 230000008569 process Effects 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000012935 Averaging Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 2
- 238000012549 training Methods 0.000 abstract description 19
- 230000004927 fusion Effects 0.000 abstract description 12
- 238000010586 diagram Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000008570 general process Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10048—Infrared image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method for fusing an infrared image and a visible light image, which comprises the following steps: (1) randomly extracting image blocks with a certain size from the 3 groups of infrared image and visible light image pairs to obtain a training image sample group, training the image sample group and obtaining a learning dictionary; (2) respectively inputting test infrared images and visible light images, calculating to obtain sparse coefficients, storing the sparse coefficients, and then combining the sparse coefficients and a learning dictionary to obtain an initial weight map; (3) decomposing the infrared and visible light images by using an average filter to obtain a high-frequency layer and a low-frequency layer of the images; (4) optimizing the initial weight map by adopting a guide filtering operation; (5) respectively fusing a high-frequency layer and a low-frequency layer of the infrared and visible light image pair according to the optimized weight graph; (6) and calculating to obtain a final fusion image according to the fused high-frequency layer and low-frequency layer. The invention has excellent effect on fusing the infrared image and the visible light image, and the fused result has abundant detail and gradient information.
Description
Technical Field
The invention belongs to the technical field of computer vision, and relates to a method for fusing an infrared image and a visible light image, in particular to the fusion of the infrared image and the visible light image in the same scene.
Background
Infrared and visible light images play an important role in transportation systems. Infrared images are obtained from the thermal radiation emitted by objects, which are less affected by weather and light conditions. However, background information in infrared images is often missing. In contrast, an image of visible light contains more texture information, however it is very susceptible to the imaging environment and lighting conditions. Infrared and visible image fusion techniques may fuse infrared and visible image pairs into one image. The fused image contains texture information from the visible image and thermal radiation details from the infrared image, thereby facilitating human observation and computer analysis. According to the different processing fields, the current main infrared and visible light image fusion methods can be divided into two types: a spatial domain method and a transform domain method. The spatial domain method directly fuses infrared and visible light image pairs through a fusion rule. A typical fusion method is to average the infrared image and the visible image. Unfortunately, this approach tends to produce fused images that are unsatisfactory. To address this problem, one decomposes the infrared and visible light image pairs into a base layer and a detail layer using a guided filter, and then fuses them using a guided filter. However, this method does not provide accurate activity level measurements. The activity level is measured from the image gradient. Some methods have proposed a method of fusing and blending multi-scale decomposed images by combining a gaussian filter with a bilateral filter, but the decomposition process of this method consumes a lot of time. Recently, some image fusion methods based on deep learning are proposed, i.e., using convolutional neural networks for image fusion. In order to improve the performance, the Laplacian pyramid is adopted to carry out multi-scale decomposition, and the self-similarity of the image is utilized to optimize the network model. Following different directions, an infrared and visible image fusion model based on generation of a countermeasure network has been proposed. Although these spatial domain methods can achieve good fusion, they also have many negative effects. They can result in overly smooth transitions at the edges of the fused image, reducing contrast, and spectral distortion of the image. For transform domain methods, multi-scale decomposition is a powerful tool that is relatively widely used, including gradient pyramids, laplacian pyramids, discrete wavelet transforms, dual-tree complex wavelet transforms, and low-pass pyramids. In addition, several geometric analysis tools are widely applied to image fusion, such as curve transformation and multi-modal image fusion, and non-downsampling contour transformation is used for decomposing a source image. Sparse representations are commonly used for machine vision tasks such as image super-resolution, face recognition and target detection. For image fusion, the general process is to compute the sparse coefficients of the source image and then fuse the coefficients through a specific rule. The latter is to convert the fusion coefficients into a result image. However, there are still some disadvantages to this type of process. Since these methods fuse coefficients in the transform domain, small changes in the coefficients may result in large changes in the spatial domain. Thus, this method suffers from severe undesirable artifacts.
Aiming at the problems, an infrared and visible light image fusion method based on sparsity and a filter is provided.
Disclosure of Invention
The present invention is directed to solving the above problems by providing an efficient infrared and visible image fusion method.
The invention realizes the purpose through the following technical scheme:
1. the infrared and visible light image fusion method comprises the following steps:
(1) randomly extracting image blocks with a certain size from the 3 groups of infrared images and visible light images to obtain a training image sample group, training the image sample group and obtaining a learning dictionary D;
(2) measuring activity level A of infrared and visible images with a trained learning dictionary DiI is (1,2), thereby obtaining a score map SiAnd i is equal to (1,2), and then the current pixel point is judged according to the score map to obtain an initial weight map Wi,i＝(1,2)；
(3) Low frequency layer LFL of image obtained by decomposing infrared and visible light images by using average filteri＝g(Ii) I ═ 1,2, and high frequency layer HFLi＝g(Ii),i＝1,2；
(4) For the initial weight map WiAnd i-1, 2 is optimized by adopting a guide filtering operation to obtain an optimized weight map
(5) According to the optimized weight map
Low frequency layer fusing infrared and visible light image pairs respectively
And a high frequency layer
(6) Combined with low-frequency layer
And a high frequency layer
A fused image can be obtained.
The basic principle of the method is as follows:
the method comprises the steps of firstly, calculating a related sparse coefficient of an input test image through a trained learning dictionary. These coefficients are used to measure the activity level of the input image pair, and then an initial weight map is derived from the activity level. The infrared and visible images are decomposed using an averaging filter to obtain a low frequency layer and a high frequency layer of the image. And optimizing the initial weight map by using a guide filter to generate an optimized weight map. And respectively fusing a low-frequency layer and a high-frequency layer of the infrared and visible light image pairs according to the optimized weight graph. And finally, combining the fused low-frequency layer and the fused high-frequency layer to obtain a fused image.
Specifically, in step (1), three infrared and visible light image pairs are used for calculating the overcomplete dictionary. A number of image patches of size n x n are sampled from the training image. By solving the following equation, an overcomplete dictionary D can be obtained:
wherein y isi∈Rn,D∈Rn×kIs a dictionary of k atoms that,
is a sparse coefficient and t is the number of non-zero elements. Then, the dictionary D is used to calculate the sparse coefficients for the input test infrared and visible light image pairs. A sliding window of size n x n is used to sample the patch pixel by pixel from the test image. Using the dictionary D, these patches can learn the coefficients by the following formula:
wherein Y isi(i ═ 1,2) is an input image patch (i.e., an infrared image or a visible image),
is a computed sparse coefficient and σ is a predefined constant.
In the step (2), the activity level A of the infrared and visible light images is measured by using the trained learning dictionary DiI ═ 1,2), and then, the sparse coefficient was calculated by the following formula
L of0Norm to measure activity level:
where the coefficients of the ith source image representing the jth patch, n is the total number of patches, and A is the computed activity map. Defining the element reshaping function as f (·), the score map SiCan be obtained using the following formula:
Si＝f(Ai),i＝(1,2)
wherein f (-) is an element reshaping function, and the following formula is utilized to obtain an initial weight graph Wi,i＝(1,2)：
W2(x,y)＝1-W1(x,y)
Wherein S is1And S2Respectively, represent a score map from a source test sample image.
In the step (3), the low-frequency layer LFL of the image is obtained by decomposing the infrared and visible light images by using an average filteri＝g(Ii) I ═ 1,2, and high frequency layer HFLi＝g(Ii) If i is 1,2 and the operator for the averaging filter operation is defined as g (·), the low-frequency layer and the high-frequency layer of the input image can be obtained by the following equations:
LFLi＝g(Ii)i＝1,2
HFLi＝Ii-LFLi i＝1,2
wherein g (-) denotes the average filtering operation, IiRepresenting a source image, i.e. an infrared image or a visible light image.
In the step (4), an initial weight map W is processediIf i ═ 1,2 is optimized by using a pilot filtering operation, and the pilot filtering operation is defined as an operator GF (·), the optimized weight map can be expressed as follows:
wherein r is1,ε1(i-1, 2) are parameters of the pilot filter,
and
optimized weight maps for the high frequency layer and the low frequency layer, respectively.
In the step (5), according to the optimized weight map
And a low frequency layer for decomposing the input image to obtain a low frequency layer and a high frequency layer to obtain a fused infrared and visible light image pair
And a high frequency layer
This process can be represented by the following equation:
wherein
It is shown that the multiplication is at the element level,
and
respectively representing the fused low frequency layer and the high frequency layer.
In the step (6), the combination is performed according to the following formula
And
a fused image F can be obtained:
the invention has the beneficial effects that:
the traditional infrared and visible light image fusion method based on multi-scale decomposition needs to decompose a source image into more than two layers. This may result in significant computational costs. The present invention only requires the decomposition of the infrared and visible light image pair into a low frequency layer and a high frequency layer. Thus, the present invention has a lower computational complexity, since it does not require re-dimensioning. Second, the present invention utilizes sparse coefficients to measure activity levels. This takes full advantage of the image characteristics and avoids large negative effects on the measured activity level caused by inaccuracies, such as ringing, occlusion and edge blurring. In addition, the guide filters are respectively used for optimizing the weight mapping of the low-frequency layer and the high-frequency layer in the invention. By doing so, the representative information in the infrared and visible light images can be well preserved. Therefore, the effect of visually fusing images can be enhanced.
Drawings
FIG. 1-1 is an overall framework diagram in an embodiment of the invention;
FIGS. 1-2 are illustrations of one of an infrared image of a training set sample in an embodiment of the present invention;
FIGS. 1-3 illustrate a second example of an infrared image of a sample in a training set in accordance with an embodiment of the present invention;
FIGS. 1-4 illustrate a third example of an infrared image of a sample in a training set in accordance with an embodiment of the present invention;
FIGS. 1-5 are one of visible light images of a training set sample in an embodiment of the present invention;
FIGS. 1-6 show a second visible light image of a training set sample in accordance with an embodiment of the present invention;
FIGS. 1-7 are three visible light images of a training set sample in accordance with an embodiment of the present invention;
FIG. 2-1 is one of the image blocks extracted from the training set samples in an embodiment of the present invention;
FIG. 2-2 is a learning dictionary D obtained by training image sample sets according to an embodiment of the present invention;
FIG. 3-1 is one of the infrared images of the test sample in an embodiment of the present invention;
FIG. 3-2 shows a second infrared image of the test sample in an embodiment of the present invention;
3-3 are three infrared images of a test sample in an embodiment of the present invention;
FIGS. 3-4 are four infrared images of a test specimen in an embodiment of the present invention;
FIGS. 3-5 are one of the visible light images of the test specimen in an embodiment of the present invention;
FIGS. 3-6 show a second visible light image of the test specimen in accordance with one embodiment of the present invention;
FIGS. 3-7 are three visible light images of a test specimen in accordance with an embodiment of the present invention;
FIGS. 3-8 are four visible light images of a test specimen in an embodiment of the present invention;
FIG. 4-1 is one of the fused images of the test sample in an embodiment of the present invention;
FIG. 4-2 shows a second fused image of a test sample according to an embodiment of the present invention;
FIGS. 4-3 illustrate a third fused image of a test sample in an embodiment of the present invention;
fig. 4-4 are four of the fused images of the test sample in the embodiment of the present invention.
Detailed Description
The invention will be further illustrated with reference to the following specific examples and the accompanying drawings:
example (b):
in order to make the fusion method of the present invention more easily understood and approximate to the real application, the whole process is described from the beginning of dictionary learning of the original training samples to the completion of image fusion, wherein the whole process includes the core fusion method of the present invention, and the general frame diagram is shown in fig. 1-1:
(1) fig. 1-2, 1-3, and 1-4 are infrared images of an original training sample, and fig. 1-5, 1-6, and 1-7 are infrared images of an original training sample. Randomly extracting image blocks with a certain size from the training sample images to obtain a sample set, in this example, setting the size of the image blocks to be 8 × 8, as shown in fig. 2-1. The image sample set is trained again and a learning dictionary D is obtained, with the dictionary size set to 64 × 512, see fig. 2-2. By solving the following equation, an overcomplete dictionary D can be obtained:
wherein y isi∈Rn,D∈Rn×kIs a dictionary of k atoms that,
is a sparse coefficient and t is the number of non-zero elements;
(2) and (3) calculating the sparse coefficient of the input test image by using the dictionary updated after training in the step (1). In the test sample sparse coding stage, we use a sliding window with the same size as the image block used in the training stage to select 4 groups of test sample infrared and visible light image pairs, such as fig. 3-1 and 3-5, fig. 3-2 and 3-6, fig. 3-3 and 3-7, and fig. 3-4 and 3-8, and sample the input image set pixel by pixel from the image pairs of the test samples, and can learn the sparse coefficients by the following formula:
wherein Y isi(i ═ 1,2) is an input image patch (i.e., an infrared image or a visible image),
is the calculated sparse coefficient, σ is a predefined constant, and σ is set to 5 in the present invention. Measuring activity level A of infrared and visible images with a trained learning dictionary DiI is (1,2), and then, the sparse coefficient obtained by the calculation in the step (2) is used as the sparse coefficient
L of0Norm to measure activity level:
where the coefficients of the ith source image representing the jth patch, n is the total number of patches, and A is the computed activity map. Defining the element reshaping function as f (·), the score map SiCan be obtained using the following formula:
Si＝f(Ai),i＝(1,2)
wherein f (-) is an element reshaping function, and the following formula is utilized to obtain an initial weight graph Wi,i＝(1,2)：
W2(x,y)＝1-W1(x,y)
Wherein S is1And S2Respectively, represent a score map from a source test sample image.
(3) Low frequency layer LFL of image obtained by decomposing 4 groups of test infrared and visible light image pairs by using average filteri＝g(Ii) I ═ 1,2, and high frequency layer HFLi＝g(Ii) If i is 1,2 and the operator for the averaging filter operation is defined as g (·), the low-frequency layer and the high-frequency layer of the input image can be obtained by the following equations:
LFLi＝g(Ii)i＝1,2
HFLi＝Ii-LFLi i＝1,2
wherein g (-) denotes the average filtering operation, IiRepresenting a source image, i.e. an infrared image or a visible light image.
(4) For the initial weight graph W obtained in the step (2)iIf i ═ 1,2 is optimized by using a pilot filtering operation, and the pilot filtering operation is defined as an operator GF (·), the optimized weight map can be expressed as follows:
wherein r is1,ε1(i ═ 1,2) is a parameter of the pilot filter, and is used in the present invention to optimize Wi HLeads the parameters r1 and epsilon of the filter1Set to 15 and 0.15, respectively, for optimizing Wi LLeads the parameters r2 and epsilon of the filter2Set to 7 and 1e-6 respectively,
and
optimized weight maps for the high frequency layer and the low frequency layer, respectively.
(5) According to the optimized weight map obtained in the step (4)
And (4) obtaining the low-frequency layer fused with the infrared and visible light image pair by the low-frequency layer and the high-frequency layer of the input test image obtained in the step (3)
And a high frequency layer
This process can be represented by the following equation:
wherein
It is shown that the multiplication is at the element level,
and
respectively representing the fused low frequency layer and the high frequency layer.
(6) Combining the obtained from step (5) according to the formula
And
a fused image F can be obtained:
and (3) comparing the 4 groups of test images with the corresponding infrared and visible light images according to the fused images obtained in the step (6) shown in fig. 4-1, 4-2, 4-3 and 4-4, wherein the fused images have more details than the infrared images and the visible light images, namely, the steps (2) to (5) are the main steps of the fusion method of the invention.
In this example, we will use a set of parameters (bi-directional information MI, edge preserving Q) that are common in the field of subjective vision and objective image processing as a measure of the amount of information involvedGAnd characteristic mutual information FMI and the like) to compare the source image with the fused image, thereby verifying the reliability of the invention.
The above embodiments are only preferred embodiments of the present invention, and are not intended to limit the technical solutions of the present invention, so long as the technical solutions can be realized on the basis of the above embodiments without creative efforts, which should be considered to fall within the protection scope of the patent of the present invention.
Claims (2)
1. The invention relates to a method for fusing an infrared image and a visible light image, which is characterized by comprising the following steps: the method comprises the following steps:
(1) measuring activity level A of infrared and visible images with a trained learning dictionary DiI is (1,2), thereby obtaining a score map SiAnd i is equal to (1,2), and then the current pixel point is judged according to the score map to obtain an initial weight map Wi,i＝(1,2)；
(2) Infrared and visible mapping using averaging filtersImage decomposition to obtain low-frequency layer LFL of imagei＝g(Ii) I ═ 1,2, and high frequency layer HFLi＝g(Ii),i＝1,2；
(3) For the initial weight map WiAnd i-1, 2 is optimized by adopting a guide filtering operation to obtain an optimized weight map
(4) According to the optimized weight map
Low frequency layer fusing infrared and visible light image pairs respectively
And a high frequency layer
2. The infrared image and visible image fusion method according to claim 1, characterized in that:
in the step (1), the activity level A of the infrared and visible light images is measured by using the trained learning dictionary DiI ═ 1,2), and then, the sparse coefficient was calculated by the following formula
L of0Norm to measure activity level:
where the coefficients of the ith source image representing the jth patch, n is the total number of patches, A is the calculated activity map, specifying an element reshaping function of f (-), score map SiCan be obtained using the following formula:
Si＝f(Ai),i＝(1,2)
wherein f (-) is an element reshaping function, and the following formula is utilized to obtain an initial weight graph Wi,i＝(1,2)：
W2(x,y)＝1-W1(x,y)
Wherein S is1And S2Respectively representing score maps from the source test sample images;
in the step (2), the low-frequency layer LFL of the image is obtained by decomposing the infrared and visible light images by using an average filteri＝g(Ii) I ═ 1,2, and high frequency layer HFLi＝g(Ii) If i is 1,2 and the operator for the averaging filter operation is defined as g (·), the low-frequency layer and the high-frequency layer of the input image can be obtained by the following equations:
LFLi＝g(Ii)i＝1,2
HFLi＝Ii-LFLii＝1,2
wherein g (-) denotes the average filtering operation, IiRepresenting a source image, i.e. an infrared image or a visible light image;
in the step (3), an initial weight map W is processediIf i ═ 1,2 is optimized by using a pilot filtering operation, and the pilot filtering operation is defined as an operator GF (·), the optimized weight map can be expressed as follows:
wherein r is1,ε1(i-1, 2) are parameters of the pilot filter,
and
respectively optimizing weight graphs of a high-frequency layer and a low-frequency layer;
in the step (4), according to the optimized weight map
Low frequency layer fusing infrared and visible light image pairs respectively
And a high frequency layer
This process can be represented by the following equation:
wherein
It is shown that the multiplication is at the element level,
and
respectively representing the fused low frequency layer and the high frequency layer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010768408.9A CN114066786A (en) | 2020-08-03 | 2020-08-03 | Infrared and visible light image fusion method based on sparsity and filter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010768408.9A CN114066786A (en) | 2020-08-03 | 2020-08-03 | Infrared and visible light image fusion method based on sparsity and filter |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114066786A true CN114066786A (en) | 2022-02-18 |
Family
ID=80231647
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010768408.9A Pending CN114066786A (en) | 2020-08-03 | 2020-08-03 | Infrared and visible light image fusion method based on sparsity and filter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114066786A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114708178A (en) * | 2022-03-30 | 2022-07-05 | 北京理工大学 | Remote sensing image fusion method based on guided filtering and sparse representation |
| CN117218048A (en) * | 2023-11-07 | 2023-12-12 | 天津市测绘院有限公司 | Infrared and visible light image fusion method based on three-layer sparse smooth model |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104021537A (en) * | 2014-06-23 | 2014-09-03 | 西北工业大学 | Infrared and visible image fusion method based on sparse representation |
| CN104809734A (en) * | 2015-05-11 | 2015-07-29 | 中国人民解放军总装备部军械技术研究所 | Infrared image and visible image fusion method based on guide filtering |
| CN106204509A (en) * | 2016-07-07 | 2016-12-07 | 西安电子科技大学 | Based on region characteristic infrared and visible light image fusion method |
| CN106981058A (en) * | 2017-03-29 | 2017-07-25 | 武汉大学 | A kind of optics based on sparse dictionary and infrared image fusion method and system |
| CN107341786A (en) * | 2017-06-20 | 2017-11-10 | 西北工业大学 | The infrared and visible light image fusion method that wavelet transformation represents with joint sparse |
| CN108052988A (en) * | 2018-01-04 | 2018-05-18 | 常州工学院 | Guiding conspicuousness image interfusion method based on wavelet transformation |
| CN108389158A (en) * | 2018-02-12 | 2018-08-10 | 河北大学 | A kind of infrared and visible light image interfusion method |
| CN108549874A (en) * | 2018-04-19 | 2018-09-18 | 广州广电运通金融电子股份有限公司 | A kind of object detection method, equipment and computer readable storage medium |
| CN109035189A (en) * | 2018-07-17 | 2018-12-18 | 桂林电子科技大学 | Infrared and weakly visible light image fusion method based on Cauchy's ambiguity function |
| CN109064437A (en) * | 2018-07-11 | 2018-12-21 | 中国人民解放军国防科技大学 | Image fusion method based on guided filtering and online dictionary learning |
| CN109447909A (en) * | 2018-09-30 | 2019-03-08 | 安徽四创电子股份有限公司 | The infrared and visible light image fusion method and system of view-based access control model conspicuousness |
| CN109919884A (en) * | 2019-01-30 | 2019-06-21 | 西北工业大学 | Infrared and visible light image fusion method based on gaussian filtering weighting |
| CN110111290A (en) * | 2019-05-07 | 2019-08-09 | 电子科技大学 | A kind of infrared and visible light image fusion method based on NSCT and structure tensor |
| CN110175970A (en) * | 2019-05-20 | 2019-08-27 | 桂林电子科技大学 | Based on the infrared and visible light image fusion method for improving FPDE and PCA |
| CN111462028A (en) * | 2020-03-16 | 2020-07-28 | 中国地质大学（武汉） | Infrared and visible light image fusion method based on phase consistency and target enhancement |
-
2020
- 2020-08-03 CN CN202010768408.9A patent/CN114066786A/en active Pending
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104021537A (en) * | 2014-06-23 | 2014-09-03 | 西北工业大学 | Infrared and visible image fusion method based on sparse representation |
| CN104809734A (en) * | 2015-05-11 | 2015-07-29 | 中国人民解放军总装备部军械技术研究所 | Infrared image and visible image fusion method based on guide filtering |
| CN106204509A (en) * | 2016-07-07 | 2016-12-07 | 西安电子科技大学 | Based on region characteristic infrared and visible light image fusion method |
| CN106981058A (en) * | 2017-03-29 | 2017-07-25 | 武汉大学 | A kind of optics based on sparse dictionary and infrared image fusion method and system |
| CN107341786A (en) * | 2017-06-20 | 2017-11-10 | 西北工业大学 | The infrared and visible light image fusion method that wavelet transformation represents with joint sparse |
| CN108052988A (en) * | 2018-01-04 | 2018-05-18 | 常州工学院 | Guiding conspicuousness image interfusion method based on wavelet transformation |
| CN108389158A (en) * | 2018-02-12 | 2018-08-10 | 河北大学 | A kind of infrared and visible light image interfusion method |
| CN108549874A (en) * | 2018-04-19 | 2018-09-18 | 广州广电运通金融电子股份有限公司 | A kind of object detection method, equipment and computer readable storage medium |
| CN109064437A (en) * | 2018-07-11 | 2018-12-21 | 中国人民解放军国防科技大学 | Image fusion method based on guided filtering and online dictionary learning |
| CN109035189A (en) * | 2018-07-17 | 2018-12-18 | 桂林电子科技大学 | Infrared and weakly visible light image fusion method based on Cauchy's ambiguity function |
| CN109447909A (en) * | 2018-09-30 | 2019-03-08 | 安徽四创电子股份有限公司 | The infrared and visible light image fusion method and system of view-based access control model conspicuousness |
| CN109919884A (en) * | 2019-01-30 | 2019-06-21 | 西北工业大学 | Infrared and visible light image fusion method based on gaussian filtering weighting |
| CN110111290A (en) * | 2019-05-07 | 2019-08-09 | 电子科技大学 | A kind of infrared and visible light image fusion method based on NSCT and structure tensor |
| CN110175970A (en) * | 2019-05-20 | 2019-08-27 | 桂林电子科技大学 | Based on the infrared and visible light image fusion method for improving FPDE and PCA |
| CN111462028A (en) * | 2020-03-16 | 2020-07-28 | 中国地质大学（武汉） | Infrared and visible light image fusion method based on phase consistency and target enhancement |
Non-Patent Citations (6)
| Title |
|---|
| JINGYUE CHEN 等: "A Novel Infrared Image Enhancement Based on Correlation Measurement of Visible Image for Urban Traffic Surveillance Systems", 《JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS》 * |
| QILEI LI 等: "Multi-Focus Image Fusion Method for Vision Sensor Systems via Dictionary Learning with Guided Filter", 《SENSORS》 * |
| YAN HUANG 等: "Infrared and Visible Image Fusion Based on Different Constraints in the Non-Subsampled Shearlet Transform Domain", 《SENSORS》 * |
| 李毅: "基于深度学习的图像融合算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王小焙: "基于引导滤波与总变分的图像融合算法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 王建: "基于改进的稀疏表示和神经网络的图像融合研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114708178A (en) * | 2022-03-30 | 2022-07-05 | 北京理工大学 | Remote sensing image fusion method based on guided filtering and sparse representation |
| CN117218048A (en) * | 2023-11-07 | 2023-12-12 | 天津市测绘院有限公司 | Infrared and visible light image fusion method based on three-layer sparse smooth model |
| CN117218048B (en) * | 2023-11-07 | 2024-03-08 | 天津市测绘院有限公司 | Infrared and visible light image fusion method based on three-layer sparse smooth model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Image dehazing using residual-based deep CNN | |
| Zhang et al. | A survey of restoration and enhancement for underwater images | |
| Xu et al. | Review of video and image defogging algorithms and related studies on image restoration and enhancement | |
| Sahu et al. | Different image fusion techniques–a critical review | |
| CN104091314B (en) | Turbulence-degraded image blind restoration method based on edge prediction and sparse ratio regular constraints | |
| CN113837974B (en) | NSST domain power equipment infrared image enhancement method based on improved BEEPS filtering algorithm | |
| Li et al. | Multifocus Image Fusion Using Wavelet‐Domain‐Based Deep CNN | |
| CN113673590A (en) | Rain removing method, system and medium based on multi-scale hourglass dense connection network | |
| Gao et al. | Improving the performance of infrared and visible image fusion based on latent low-rank representation nested with rolling guided image filtering | |
| Ding et al. | U 2 D 2 Net: Unsupervised unified image dehazing and denoising network for single hazy image enhancement | |
| Luo et al. | Infrared and visible image fusion based on visibility enhancement and hybrid multiscale decomposition | |
| Sahu et al. | Trends and prospects of techniques for haze removal from degraded images: A survey | |
| CN114066786A (en) | Infrared and visible light image fusion method based on sparsity and filter | |
| Chen et al. | The enhancement of catenary image with low visibility based on multi-feature fusion network in railway industry | |
| CN114170103A (en) | Electrical equipment infrared image enhancement method | |
| Chang | Single underwater image restoration based on adaptive transmission fusion | |
| Fu et al. | An anisotropic Gaussian filtering model for image de-hazing | |
| Lei et al. | Low-light image enhancement using the cell vibration model | |
| Kaur et al. | Automated Multimodal image fusion for brain tumor detection | |
| CN111461999A (en) | SAR image speckle suppression method based on super-pixel similarity measurement | |
| CN110880192A (en) | Image DCT coefficient distribution fitting method based on probability density function dictionary | |
| CN113781375B (en) | Vehicle-mounted vision enhancement method based on multi-exposure fusion | |
| Guan et al. | DiffWater: Underwater Image Enhancement Based on Conditional Denoising Diffusion Probabilistic Model | |
| Huang et al. | Attention-based for multiscale fusion underwater image enhancement | |
| CN116309221A (en) | Method for constructing multispectral image fusion model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20220218 |